Microglial activation and atrophy in frontal cortex predict executive dysfunction in frontotemporal dementia
AbstractBackgroundFrontotemporal dementia is clinically and neuropathologically heterogeneous, but atrophy, neuroinflammation, and executive dysfunction occur in each of the principal variants. Across the clinical spectrum of frontotemporal dementia, we assessed the predictive value of in vivo neuroimaging measures of grey‐matter volume (from structural MRI) and microglial activation (from [11C]PK11195 PET) on the rate of future executive decline. We hypothesised a detrimental effect of inflammation and atrophy severity on executive dysfunction progression.MethodThirty patients with frontotemporal dementia underwent a baseline multi‐modal imaging assessment, including [11C]PK11195 PET to index microglial activation and structural MRI for atrophy, as part of the Neuroimaging of Inflammation in Memory and Related Other Disorders (NIMROD) study. Cognitive impairments were assessed at baseline and serially for up to 5 years with the revised Addenbrooke's Cognitive Examination (ACE‐R). Regional grey‐matter volumes and [11C]PK11195 binding potentials were averaged in four regions of interest: left and right frontal and temporal lobes. Linear mixed models were applied to the longitudinal ACE‐R attention/executive sub‐score. Regional CSF‐corrected PET values and TIV‐corrected grey‐matter volumes were included in linear mixed models as predictors, with age and education as covariates. Specifically, negative associations of atrophy and inflammation with annual rate of executive decline were tested.ResultPatients showed a mean 3.3‐point loss per year on the ACE‐R attention/executive sub‐score (p<0.001, Figure 1). Faster decline in ACE‐R was associated with reduced baseline grey‐matter volume of the left frontal lobe (beta=1.000, SE=0.285, p=0.000789, p‐FDR= 0.00316; Figure 2) and increased inflammation in frontal regions bilaterally (Left: beta=‐0.722, SE=0.251, p=0.00528, p‐FDR=0.0105, Figure 2; Right: Estimate=‐0.623, SE=0.256, p=0.0175, p‐FDR=0.0233). These associations remain significant when including both left frontal grey‐matter volume and [11C]PK11195 binding as predictors in the same model, and after controlling for baseline attention/executive scores. In these regions, grey‐matter volumes and inflammation levels were negatively correlated (left: r=‐0.414, p<0.001; right: r=‐0.374, p<0.001).ConclusionImaging markers for atrophy and microglial activation provide useful and independent information to evaluate and stratify patients with frontotemporal dementia. This may improve cohort selection in clinical trials and highlights the potential for immunomodulatory treatment strategies in frontotemporal dementia.